Re-enabled overwriting of URL field (#1237)

Enabled overwriting of URL field

URL_FIELD_NAME is usually used as self-link in links.
However it should be allowed to be overwritten as long
as not HyperlinkedIdentifyField has been used.

Author: sliverc

CHANGELOG.md

@@ -8,6 +8,12 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 Note that in line with [Django REST framework policy](https://www.django-rest-framework.org/topics/release-notes/),
 any parts of the framework not mentioned in the documentation should generally be considered private API, and may be subject to change.
 
+## [Unreleased]
+
+### Fixed
+
+* Re-enabled overwriting of url field (regression since 7.0.0)
+
 ## [7.0.1] - 2024-06-06
 
 ### Added

rest_framework_json_api/renderers.py

@@ -75,13 +75,11 @@ def extract_attributes(cls, fields, resource):
         and relationships are not returned.
         """
 
-        invalid_fields = {"id", api_settings.URL_FIELD_NAME}
-
         return {
             format_field_name(field_name): value
             for field_name, value in resource.items()
             if field_name in fields
-            and field_name not in invalid_fields
+            and field_name != "id"
             and not is_relationship_field(fields[field_name])
         }
 
@@ -449,7 +447,10 @@ def _filter_sparse_fields(cls, serializer, fields, resource_name):
                     if field.field_name in sparse_fields
                     # URL field is not considered a field in JSON:API spec
                     # but a link so need to keep it
-                    or field.field_name == api_settings.URL_FIELD_NAME
+                    or (
+                        field.field_name == api_settings.URL_FIELD_NAME
+                        and isinstance(field, relations.HyperlinkedIdentityField)
+                    )
                 }
 
         return fields
@@ -486,7 +487,7 @@ def build_json_resource_obj(
             resource_data["relationships"] = relationships
         # Add 'self' link if field is present and valid
         if api_settings.URL_FIELD_NAME in resource and isinstance(
-            fields[api_settings.URL_FIELD_NAME], relations.RelatedField
+            fields[api_settings.URL_FIELD_NAME], relations.HyperlinkedIdentityField
         ):
             resource_data["links"] = {"self": resource[api_settings.URL_FIELD_NAME]}
 

rest_framework_json_api/serializers.py

@@ -6,6 +6,7 @@
 from django.utils.module_loading import import_string as import_class_from_dotted_path
 from django.utils.translation import gettext_lazy as _
 from rest_framework.exceptions import ParseError
+from rest_framework.relations import HyperlinkedIdentityField
 
 # star import defined so `rest_framework_json_api.serializers` can be
 # a simple drop in for `rest_framework.serializers`
@@ -94,9 +95,12 @@ def _readable_fields(self):
                         field
                         for field in readable_fields
                         if field.field_name in sparse_fields
-                        # URL field is not considered a field in JSON:API spec
-                        # but a link so need to keep it
-                        or field.field_name == api_settings.URL_FIELD_NAME
+                        # URL_FIELD_NAME is the field used as self-link to resource
+                        # however only when it is a HyperlinkedIdentityField
+                        or (
+                            field.field_name == api_settings.URL_FIELD_NAME
+                            and isinstance(field, HyperlinkedIdentityField)
+                        )
                         # ID is a required field which might have been overwritten
                         # so need to keep it
                         or field.field_name == "id"

tests/conftest.py

@@ -8,6 +8,7 @@
     ManyToManySource,
     ManyToManyTarget,
     NestedRelatedSource,
+    URLModel,
 )
 
 
@@ -36,6 +37,11 @@ def model(db):
     return BasicModel.objects.create(text="Model")
 
 
+@pytest.fixture
+def url_instance(db):
+    return URLModel.objects.create(text="Url", url="https://example.com")
+
+
 @pytest.fixture
 def foreign_key_target(db):
     return ForeignKeyTarget.objects.create(name="Target")

tests/models.py

@@ -18,6 +18,14 @@ class Meta:
         ordering = ("id",)
 
 
+class URLModel(DJAModel):
+    url = models.URLField()
+    text = models.CharField(max_length=100)
+
+    class Meta:
+        ordering = ("id",)
+
+
 # Models for relations tests
 # ManyToMany
 class ManyToManyTarget(DJAModel):

tests/serializers.py

@@ -8,6 +8,7 @@
     ManyToManySource,
     ManyToManyTarget,
     NestedRelatedSource,
+    URLModel,
 )
 
 
@@ -17,6 +18,15 @@ class Meta:
         model = BasicModel
 
 
+class URLModelSerializer(serializers.ModelSerializer):
+    class Meta:
+        fields = (
+            "text",
+            "url",
+        )
+        model = URLModel
+
+
 class ForeignKeyTargetSerializer(serializers.ModelSerializer):
     class Meta:
         fields = ("name",)

tests/test_views.py

@@ -21,6 +21,7 @@
     ForeignKeyTargetViewSet,
     ManyToManySourceViewSet,
     NestedRelatedSourceViewSet,
+    URLModelViewSet,
 )
 
 
@@ -182,6 +183,42 @@ def test_list_with_default_included_resources(self, client, foreign_key_source):
             }
         ] == result["included"]
 
+    @pytest.mark.urls(__name__)
+    def test_list_allow_overwriting_url_field(self, client, url_instance):
+        """
+        Test overwriting of url is possible.
+
+        URL_FIELD_NAME which is set to 'url' per default is used as self in links.
+        However if field is overwritten and not a HyperlinkedIdentityField it should be allowed
+        to use as a attribute as well.
+        """
+
+        url = reverse("urlmodel-list")
+        response = client.get(url)
+        assert response.status_code == status.HTTP_200_OK
+        data = response.json()["data"]
+        assert data == [
+            {
+                "type": "URLModel",
+                "id": str(url_instance.pk),
+                "attributes": {"text": "Url", "url": "https://example.com"},
+            }
+        ]
+
+    @pytest.mark.urls(__name__)
+    def test_list_allow_overwiritng_url_with_sparse_fields(self, client, url_instance):
+        url = reverse("urlmodel-list")
+        response = client.get(url, data={"fields[URLModel]": "text"})
+        assert response.status_code == status.HTTP_200_OK
+        data = response.json()["data"]
+        assert data == [
+            {
+                "type": "URLModel",
+                "id": str(url_instance.pk),
+                "attributes": {"text": "Url"},
+            }
+        ]
+
     @pytest.mark.urls(__name__)
     def test_retrieve(self, client, model):
         url = reverse("basic-model-detail", kwargs={"pk": model.pk})
@@ -495,6 +532,7 @@ def patch(self, request, *args, **kwargs):
 # configuration in general
 router = SimpleRouter()
 router.register(r"basic_models", BasicModelViewSet, basename="basic-model")
+router.register(r"url_models", URLModelViewSet)
 router.register(r"foreign_key_sources", ForeignKeySourceViewSet)
 router.register(r"foreign_key_targets", ForeignKeyTargetViewSet)
 router.register(

tests/views.py

@@ -5,6 +5,7 @@
     ForeignKeyTarget,
     ManyToManySource,
     NestedRelatedSource,
+    URLModel,
 )
 from tests.serializers import (
     BasicModelSerializer,
@@ -13,6 +14,7 @@
     ForeignKeyTargetSerializer,
     ManyToManySourceSerializer,
     NestedRelatedSourceSerializer,
+    URLModelSerializer,
 )
 
 
@@ -22,6 +24,12 @@ class BasicModelViewSet(ModelViewSet):
     ordering = ["text"]
 
 
+class URLModelViewSet(ModelViewSet):
+    serializer_class = URLModelSerializer
+    queryset = URLModel.objects.all()
+    ordering = ["url"]
+
+
 class ForeignKeySourceViewSet(ModelViewSet):
     serializer_class = ForeignKeySourceSerializer
     queryset = ForeignKeySource.objects.all()